Entangled Bimodal Vision in Vehicles for Decision During Risk Situation

On the foundation of a cost-effective embedded system, this work integrates the principles of entanglement and learned simulated flows in a vehicular vision for situational awareness that can be integrated with a vehicular cloud for decision making. The road scenes of bimodal vision are extracted, and the vehicles are detected according to zoned-based on locations. The attributes of the vehicle along with the depth information are extracted and analyzed for recognition purposes. The system employs two cameras that are used for decision making. The decision-making attributes are weighted and entangled to optimize the decision process. Ten real road situations are modelled, using the finite element method, which are trained and integrated with the entanglement decision making. The proposed method implementation results show low RMSE in risk prediction in comparison to the monocular vision system and conventional fusion of multi-modal vision vehicular vision. The system implementation results show that the system is effective and promising for the above mentioned road conditions.